INDEX

Explanations

mentions of knowledge or familiarity in a context where information is being shared or referenced

oai_token-act-pair · gpt-3.5-turbo

references to prior knowledge or familiarity with specific content

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ickey

-0.67

inals

-0.65

 envelope

-0.63

Els

-0.62

jri

-0.62

ella

-0.59

ospel

-0.59

rame

-0.57

 ransom

-0.57

 subsystem

-0.57

POSITIVE LOGITS

mentioned

0.82

taboola

0.79

 WATCHED

0.74

Ü

0.72

Ago

0.66

ebin

0.66

TPPStreamerBot

0.65

noticed

0.64

kens

0.63

ITNESS

0.62

Activations Density 0.228%

mentions of knowledge or familiarity in a context where information is being shared or referenced

references to prior knowledge or familiarity with specific content

No Comments

No Known Activations