INDEX

Explanations

instances of the word "policy"

oai_token-act-pair · gpt-3.5-turbo

references to various policies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ITNESS

-0.90

htaking

-0.84

ingly

-0.80

nces

-0.80

 Flavoring

-0.77

apy

-0.72

bane

-0.72

Ago

-0.71

 Stain

-0.71

ymes

-0.70

POSITIVE LOGITS

 enforcement

0.96

holders

0.90

 Enforcement

0.89

holder

0.86

 makers

0.84

 prohibiting

0.83

 policy

0.83

 enforced

0.78

 abiding

0.76

 policies

0.75

Activations Density 0.027%

instances of the word "policy"

references to various policies

No Comments

No Known Activations