INDEX

Explanations

phrases related to evaluations or judgments of people or actions, particularly focused on criticizing or expressing negative opinions

oai_token-act-pair · gpt-3.5-turbo

evaluations of a person's character or performance

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inav

-0.86

artments

-0.77

 partitions

-0.76

urry

-0.74

iners

-0.74

osate

-0.73

 spans

-0.73

imeters

-0.71

 Locations

-0.71

Wars

-0.70

POSITIVE LOGITS

 assassinated

1.10

 persona

1.09

 willing

1.08

 incapable

1.04

 obsessed

1.03

 happiest

1.02

 himself

1.02

 pissed

0.99

 afraid

0.99

 capable

0.98

Activations Density 0.467%

phrases related to evaluations or judgments of people or actions, particularly focused on criticizing or expressing negative opinions

evaluations of a person's character or performance

No Comments

No Known Activations