INDEX

Explanations

words related to protection or defense against different kinds of threats or dangers

oai_token-act-pair · gpt-3.5-turbo

references to threats or dangers that require protection or defense

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soDeliveryDate

-0.94

Austral

-0.78

actionDate

-0.72

ACTED

-0.72

Phones

-0.70

Pie

-0.66

Circ

-0.65

VK

-0.65

çĶ°

-0.65

jong

-0.65

POSITIVE LOGITS

 wrath

0.92

 pressures

0.88

 pitfalls

0.87

 scrutiny

0.85

 onslaught

0.82

 toxicity

0.78

 hardships

0.75

 contaminants

0.74

 predators

0.74

 allerg

0.74

Activations Density 0.524%

words related to protection or defense against different kinds of threats or dangers

references to threats or dangers that require protection or defense

No Comments

No Known Activations