INDEX

Explanations

phrases related to communication, interaction, and information sharing

oai_token-act-pair · gpt-3.5-turbo

references to collective activities or interactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

soever

-0.67

stroke

-0.65

alky

-0.65

pee

-0.64

thinkable

-0.63

uster

-0.61

netflix

-0.61

 Bundy

-0.60

 alienation

-0.60

Pain

-0.60

POSITIVE LOGITS

 ourselves

1.05

bsite

0.92

 Spons

0.88

 previews

0.79

our

0.78

blogs

0.78

 publish

0.75

 ende

0.75

 recommend

0.72

 invite

0.70

Activations Density 0.313%

phrases related to communication, interaction, and information sharing

references to collective activities or interactions

No Comments

No Known Activations