INDEX

Explanations

references to political figures and countries

oai_token-act-pair · gpt-3.5-turbo

references to numeric values and measurements associated with technology

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

 tranquil

-0.75

heit

-0.65

ĸļ

-0.64

Quantity

-0.62

URR

-0.62

cla

-0.62

 Cyrus

-0.61

 mistrust

-0.59

utsche

-0.59

 saddened

-0.58

POSITIVE LOGITS

iev

0.82

peat

0.77

ateral

0.76

ilver

0.76

idon

0.72

GHz

0.71

OD

0.71

handled

0.70

thodox

0.69

teness

0.69

Activations Density 0.302%

references to political figures and countries

references to numeric values and measurements associated with technology

No Comments

No Known Activations