INDEX

Explanations

phrases expressing contrasting ideas

oai_token-act-pair · gpt-3.5-turbo

expressions of skepticism or doubt about common beliefs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iphate

-0.84

ricted

-0.76

 Inventory

-0.75

ourning

-0.74

Profile

-0.71

ctors

-0.69

 actionGroup

-0.69

hibited

-0.68

ievers

-0.67

essor

-0.64

POSITIVE LOGITS

 unavoidable

0.99

 nonetheless

0.94

 achievable

0.92

 understandable

0.89

 borne

0.88

 happening

0.86

 indicative

0.86

 sympt

0.86

 exacerbated

0.85

 consequential

0.83

Activations Density 0.395%

phrases expressing contrasting ideas

expressions of skepticism or doubt about common beliefs

No Comments

No Known Activations