INDEX

Explanations

mentions of skills or abilities

oai_token-act-pair · gpt-3.5-turbo

references to various types of skills and competencies

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

romeda

-0.89

rarily

-0.66

 Destination

-0.65

iversary

-0.64

etheless

-0.64

 ãĤµ

-0.63

rary

-0.63

Mars

-0.63

lique

-0.62

 sidel

-0.62

POSITIVE LOGITS

cape

0.98

heet

0.93

mith

0.90

 learners

0.85

ets

0.84

hooting

0.83

linger

0.82

 skills

0.82

 training

0.81

 tutor

0.77

Activations Density 0.032%

mentions of skills or abilities

references to various types of skills and competencies

No Comments

No Known Activations