INDEX

Explanations

terms related to various kinds of effects, particularly those with negative connotations

oai_token-act-pair · gpt-3.5-turbo

concepts relating to various types of effects, particularly in a negative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ilo

-0.68

bledon

-0.67

Sec

-0.65

annie

-0.62

IFA

-0.62

sted

-0.62

furt

-0.61

Bay

-0.60

illet

-0.59

imeter

-0.59

POSITIVE LOGITS

 effects

3.62

effects

3.09

 Effects

2.94

Effects

2.57

 effect

2.47

 impacts

1.98

 Effect

1.88

 consequences

1.86

Effect

1.73

effect

1.68

Activations Density 0.021%

terms related to various kinds of effects, particularly those with negative connotations

concepts relating to various types of effects, particularly in a negative context

No Comments

No Known Activations