INDEX

Explanations

phrases related to patterns or systematic processes

oai_token-act-pair · gpt-3.5-turbo

references to recognizable patterns in various contexts

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

ascular

-0.82

omez

-0.80

UGE

-0.71

UGH

-0.68

ULTS

-0.67

rican

-0.66

phies

-0.66

orean

-0.66

venture

-0.65

avez

-0.64

POSITIVE LOGITS

 pattern

0.89

pattern

0.87

gradient

0.86

 Pattern

0.84

 patterns

0.83

Pattern

0.83

 Patterns

0.78

 repeats

0.78

eering

0.76

atile

0.73

Activations Density 0.031%

phrases related to patterns or systematic processes

references to recognizable patterns in various contexts

No Comments

No Known Activations