INDEX

Explanations

adjectives related to the level of difficulty or effort required for a task

oai_token-act-pair · gpt-3.5-turbo

references to challenges or difficulties characterized by the word "easy."

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

inating

-0.66

ensional

-0.66

inated

-0.65

)].

-0.65

aina

-0.64

————————————————

-0.63

oola

-0.60

len

-0.60

ortment

-0.60

span

-0.60

POSITIVE LOGITS

 anymore

0.90

nor

0.82

 chore

0.77

achable

0.69

 Canaver

0.69

yet

0.68

 discipl

0.67

 Sasuke

0.66

È

0.66

cube

0.66

Activations Density 0.064%

adjectives related to the level of difficulty or effort required for a task

references to challenges or difficulties characterized by the word "easy."

No Comments

No Known Activations