INDEX

Explanations

phrases that include the word "sort" in reference to categorization or comparison

oai_token-act-pair · gpt-3.5-turbo

phrases that express classification or categorization

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

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

interrupted

-0.63

 Madison

-0.62

PLIC

-0.61

PER

-0.60

VICE

-0.60

 Dent

-0.59

PLE

-0.58

INT

-0.57

NZ

-0.56

VIS

-0.56

POSITIVE LOGITS

ilege

0.88

0.84

ies

0.83

ie

0.82

ative

0.81

liness

0.81

olith

0.79

iple

0.77

able

0.77

entially

0.76

Activations Density 0.033%

phrases that include the word "sort" in reference to categorization or comparison

phrases that express classification or categorization

No Comments

No Known Activations