INDEX

Explanations

specific references to a particular concept or entity

oai_token-act-pair · gpt-3.5-turbo

instances of the word "this" in relation to various contexts and topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

icons

-0.88

lee

-0.79

å§«

-0.77

okers

-0.76

acers

-0.76

mates

-0.75

visors

-0.75

ashes

-0.74

masters

-0.74

master

-0.73

POSITIVE LOGITS

 particular

1.07

 newfound

1.03

 phenomenon

1.02

 trope

1.02

 century

0.93

 avenue

0.93

 kind

0.93

 hemisphere

0.92

 millenn

0.91

 enigmatic

0.91

Activations Density 0.226%

specific references to a particular concept or entity

instances of the word "this" in relation to various contexts and topics

No Comments

No Known Activations