INDEX

Explanations

instances of specific values or parameters in code

oai_token-act-pair · gpt-3.5-turbo

instances of the word "value" with varying numerical significance commonly found in programming contexts

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

ooks

-0.77

ocamp

-0.76

ility

-0.71

kefeller

-0.70

osaurus

-0.70

dar

-0.69

alyst

-0.68

udeb

-0.68

alm

-0.67

 Citiz

-0.65

POSITIVE LOGITS

 value

0.83

 values

0.83

values

0.74

EStream

0.73

 Value

0.69

VALUE

0.67

 Values

0.65

lda

0.64

 Vance

0.64

 proposition

0.63

Activations Density 0.019%

instances of specific values or parameters in code

instances of the word "value" with varying numerical significance commonly found in programming contexts

No Comments

No Known Activations