INDEX

Explanations

terms related to manipulation or manipulative actions

oai_token-act-pair · gpt-3.5-turbo

terms related to manipulation and control

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

çĦ

-0.82

hood

-0.77

ness

-0.73

Republic

-0.71

zl

-0.70

fighter

-0.68

gap

-0.68

 Blessed

-0.67

Ath

-0.65

senal

-0.65

POSITIVE LOGITS

 manipulate

0.98

 manip

0.97

 manipulating

0.89

eering

0.87

 manipulated

0.87

 manipulation

0.86

glers

0.81

 levers

0.81

ulators

0.80

ulations

0.78

Activations Density 0.021%

terms related to manipulation or manipulative actions

terms related to manipulation and control

No Comments

No Known Activations