INDEX

Explanations

descriptions or accounts of experiences

oai_token-act-pair · gpt-3.5-turbo

instances of the word "described."

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

cot

-0.66

aghetti

-0.66

alos

-0.64

iasm

-0.64

Bus

-0.64

think

-0.63

ificial

-0.63

ammy

-0.62

ffic

-0.59

isdom

-0.59

POSITIVE LOGITS

 descriptions

0.78

urated

0.77

 symptoms

0.76

 markings

0.71

uron

0.69

details

0.69

REDACTED

0.69

urally

0.69

 aloud

0.65

ribing

0.65

Activations Density 0.029%

descriptions or accounts of experiences

instances of the word "described."

No Comments

No Known Activations