INDEX

Explanations

phrases related to observing or witnessing events or behaviors

oai_token-act-pair · gpt-3.5-turbo

instances of the word "observe" and variations thereof

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

utm

-0.76

nec

-0.75

gur

-0.75

venge

-0.74

recy

-0.71

xit

-0.71

cit

-0.69

onz

-0.69

rax

-0.69

ikuman

-0.69

POSITIVE LOGITS

 observation

0.92

 observing

0.81

 observe

0.81

 Observ

0.80

 observed

0.79

 Patterns

0.76

 patterns

0.73

eering

0.73

 observations

0.71

áµ

0.71

Activations Density 0.015%

phrases related to observing or witnessing events or behaviors

instances of the word "observe" and variations thereof

No Comments

No Known Activations