INDEX

Explanations

words related to government agencies

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the term "agency."

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

lihood

-0.91

Arn

-0.72

yle

-0.69

Klu

-0.67

kai

-0.66

Tea

-0.65

roleum

-0.63

isers

-0.63

aston

-0.63

ening

-0.63

POSITIVE LOGITS

 agencies

0.93

 tasked

0.86

 agency

0.84

ENCY

0.84

rador

0.84

ality

0.77

ional

0.76

encies

0.76

 bureaucracy

0.75

acco

0.73

Activations Density 0.021%

words related to government agencies

repeated mentions of the term "agency."

No Comments

No Known Activations