INDEX

Explanations

words related to assistive technology and terms related to affirmative or preventive actions

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

flies

-0.85

puter

-0.85

olicited

-0.84

jri

-0.82

ortium

-0.79

ufact

-0.78

ulhu

-0.78

igrate

-0.77

under

-0.73

etter

-0.73

POSITIVE LOGITS

 measures

0.96

 Measures

0.89

 minded

0.86

 tendencies

0.77

 clauses

0.77

 action

0.76

 force

0.76

 braking

0.75

 pronoun

0.74

 pronouns

0.74

Activations Density 0.131%

words related to assistive technology and terms related to affirmative or preventive actions

No Comments

No Known Activations