INDEX

Explanations

programming code snippets containing specific keywords

oai_token-act-pair · gpt-3.5-turbo

programming syntax and elements, specifically related to JavaScript functions and error handling

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ppa

-0.96

 Bennett

-0.82

Plex

-0.77

BuyableInstoreAndOnline

-0.77

tek

-0.76

gart

-0.76

 Gomez

-0.76

ĸļ

-0.75

edin

-0.74

Sau

-0.73

POSITIVE LOGITS

 function

1.34

 expressions

1.32

 operator

1.32

 predicate

1.29

 loops

1.27

expression

1.27

 conditional

1.27

 parentheses

1.26

 expression

1.23

 operators

1.21

Activations Density 0.199%

programming code snippets containing specific keywords

programming syntax and elements, specifically related to JavaScript functions and error handling

No Comments

No Known Activations