INDEX

Explanations

phrases indicating a strong level of commitment or effort towards achieving a goal

oai_token-act-pair · gpt-3.5-turbo

references to making efforts or taking actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

opened

-0.70

 Flavoring

-0.69

Ent

-0.69

aten

-0.68

Constructed

-0.68

Tru

-0.67

ipel

-0.63

Cth

-0.63

 Returning

-0.62

 Bound

-0.61

POSITIVE LOGITS

 differently

0.93

 grunt

0.76

 offline

0.72

 deed

0.68

 administr

0.67

 ashore

0.67

 unconsciously

0.67

 injustice

0.66

 wrong

0.65

 unilaterally

0.64

Activations Density 0.054%

phrases indicating a strong level of commitment or effort towards achieving a goal

references to making efforts or taking actions

No Comments

No Known Activations