INDEX

Explanations

action words related to being lost or stuck

oai_token-act-pair · gpt-3.5-turbo

occurrences of the words "lost" and "stuck" in various 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

abiding

-0.65

Ng

-0.64

chief

-0.61

ighth

-0.61

audi

-0.59

WB

-0.59

integ

-0.59

Method

-0.59

Break

-0.58

Hong

-0.56

POSITIVE LOGITS

retty

0.93

ãĤ¼

0.88

ocobo

0.76

*/(

0.71

cair

0.69

ãĤ¶

0.68

adeon

0.67

 destro

0.65

itta

0.65

chy

0.64

Activations Density 0.106%

action words related to being lost or stuck

occurrences of the words "lost" and "stuck" in various contexts

No Comments

No Known Activations