INDEX

Explanations

instances where a difficulty, hindrance, or inability is conveyed

oai_token-act-pair · gpt-3.5-turbo

negations or expressions of inability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Called

-0.74

interstitial

-0.69

amy

-0.67

 Gleaming

-0.65

 Matters

-0.64

soDeliveryDate

-0.63

heads

-0.63

 Appears

-0.62

Vert

-0.61

 torches

-0.60

POSITIVE LOGITS

 afford

0.99

 foresee

0.95

 adequately

0.83

 muster

0.81

 predict

0.80

 convince

0.80

 envision

0.80

 conceive

0.80

successfully

0.77

rehend

0.77

Activations Density 0.087%

instances where a difficulty, hindrance, or inability is conveyed

negations or expressions of inability

No Comments

No Known Activations