INDEX

Explanations

phrases related to ratings or evaluations

oai_token-act-pair · gpt-3.5-turbo

phrases related to rating or evaluating items

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

aires

-0.69

aire

-0.68

ELS

-0.67

ancies

-0.63

ciating

-0.61

 Gazette

-0.59

thr

-0.58

ORE

-0.58

externalActionCode

-0.58

 outnumbered

-0.57

POSITIVE LOGITS

this

1.43

 this

1.20

This

1.13

 THIS

1.09

 This

1.04

THIS

0.94

these

0.94

 these

0.75

past

0.69

that

0.69

Activations Density 0.246%

phrases related to ratings or evaluations

phrases related to rating or evaluating items

No Comments

No Known Activations