INDEX

Explanations

references to specific years

oai_token-act-pair · gpt-3.5-turbo

temporal markers and references to past experiences or important years

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 salute

-0.79

isode

-0.69

 Ending

-0.69

onyms

-0.67

 Unlock

-0.65

oward

-0.64

MEN

-0.64

Released

-0.62

Happy

-0.61

 Bliss

-0.61

POSITIVE LOGITS

 constraints

0.93

 inexper

0.93

 pressures

0.77

 uncertainties

0.76

 deadlines

0.75

 inexperienced

0.74

 hindsight

0.73

 constrained

0.72

 foreseeable

0.71

 schedules

0.71

Activations Density 1.514%

references to specific years

temporal markers and references to past experiences or important years

No Comments

No Known Activations