INDEX

Explanations

words related to question and answer sessions

oai_token-act-pair · gpt-3.5-turbo

references to question and answer sessions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iris

-0.77

esm

-0.73

most

-0.72

ports

-0.70

stead

-0.69

acements

-0.69

metics

-0.69

 Penal

-0.68

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.68

wikipedia

-0.68

POSITIVE LOGITS

 session

1.47

 sessions

1.30

 Session

1.12

 conversation

1.07

 interview

1.04

 discussion

1.04

 moder

0.96

 podcast

0.92

 forum

0.92

 segment

0.90

Activations Density 0.138%

words related to question and answer sessions

references to question and answer sessions

No Comments

No Known Activations