INDEX

Explanations

the word "ga" appearing at different levels of activation

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "ga"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

chard

-0.79

tz

-0.79

itaire

-0.72

icable

-0.70

 Cosponsors

-0.69

ecause

-0.69

sbm

-0.69

itarian

-0.68

ality

-0.67

alities

-0.67

POSITIVE LOGITS

terday

0.98

ignt

0.72

enaries

0.71

 Nieto

0.67

 indu

0.64

arde

0.64

andise

0.61

 Varg

0.61

veyard

0.61

udo

0.60

Activations Density 0.041%

the word "ga" appearing at different levels of activation

repeated mentions of the word "ga"

No Comments

No Known Activations