INDEX

Explanations

references to historical events or contexts

oai_token-act-pair · gpt-3.5-turbo

references to historical context and events

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

lain

-1.12

nery

-0.87

ertodd

-0.84

liga

-0.83

geon

-0.81

nel

-0.79

chy

-0.78

rays

-0.78

wich

-0.77

ldon

-0.77

POSITIVE LOGITS

 significance

0.94

orical

0.94

 revision

0.91

orically

0.91

 inaccur

0.84

 context

0.84

 inacc

0.84

 preservation

0.84

 precedent

0.81

 heritage

0.80

Activations Density 0.019%

references to historical events or contexts

references to historical context and events

No Comments

No Known Activations