INDEX

Explanations

references to the Taliban

oai_token-act-pair · gpt-3.5-turbo

mentions of the Taliban

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

abet

-0.71

yss

-0.71

secut

-0.71

isner

-0.69

efer

-0.68

gas

-0.68

++++++++++++++++

-0.67

leon

-0.66

OHN

-0.66

 Quadro

-0.65

POSITIVE LOGITS

 insurgents

1.17

istani

1.14

 insurgency

1.04

 Taliban

1.00

istan

0.97

 insurgent

0.90

 Afghan

0.82

 Haram

0.81

 militants

0.80

 infiltr

0.78

Activations Density 0.007%

references to the Taliban

mentions of the Taliban

No Comments

No Known Activations