INDEX

Explanations

phrases indicating location or containment within a particular context

oai_token-act-pair · gpt-3.5-turbo

references to context or environments surrounding various topics

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

lace

-0.69

ãĥ£

-0.69

ãĥĹ

-0.69

enez

-0.69

ãĤ³

-0.65

lan

-0.65

ple

-0.65

ãĥĵ

-0.65

 goodbye

-0.65

ãĤ·ãĥ£

-0.64

POSITIVE LOGITS

imore

0.87

¥ŀ

0.82

isine

0.81

ieth

0.78

 parentheses

0.74

animate

0.73

ciating

0.71

izabeth

0.71

isode

0.68

ĨĴ

0.68

Activations Density 0.023%

phrases indicating location or containment within a particular context

references to context or environments surrounding various topics

No Comments

No Known Activations