INDEX

Explanations

mentions of people's names and their related actions or characteristics

oai_token-act-pair · gpt-3.5-turbo

events and actions occurring in a narrative context

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

LCS

-0.77

inventoryQuantity

-0.74

etheless

-0.74

olid

-0.73

byss

-0.67

 Streaming

-0.67

GSL

-0.66

ï¸ı

-0.66

UFC

-0.65

WCS

-0.65

POSITIVE LOGITS

 startled

0.82

 astonished

0.79

 exclaim

0.78

 horrified

0.77

 exclaimed

0.75

"'

0.75

 classmate

0.74

 innocuous

0.74

 arous

0.72

 classmates

0.72

Activations Density 1.535%

mentions of people's names and their related actions or characteristics

events and actions occurring in a narrative context

No Comments

No Known Activations