INDEX

Explanations

phrases or terms indicating a range or variety of subjects or topics

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a range or variety of topics or conditions

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

mit

-0.79

imposed

-0.75

spot

-0.74

clusions

-0.70

illin

-0.70

ÄŁ

-0.69

driving

-0.69

rafted

-0.67

template

-0.67

yles

-0.66

POSITIVE LOGITS

 ranging

0.91

nesota

0.65

 unlaw

0.65

 Luxem

0.64

ãĤ¤ãĥĪ

0.63

ranging

0.62

 vari

0.62

ĸļ

0.61

ogyn

0.61

 upwards

0.61

Activations Density 0.020%

phrases or terms indicating a range or variety of subjects or topics

phrases indicating a range or variety of topics or conditions

No Comments

No Known Activations